2019-08-29 wg on euro rfr meeting item 3 update by ......aug 29, 2019 · sub group 2a: status...

Item 3: Update by subgroup 2

on term rates methodologies

Working group on euro risk-free rates, 29 August 2019

Sub Group 2A: Status update EURRFR Working Group Meeting 29.08.2019

Frankfurt am Main

Sub Group 2A Status update

2

Credit spread methodologies: Analysis of the various methodologies to calculate a fixed spread, representing the difference between EURIBOR and the €STR-based term structure methodologies

final alignment 1.

Backward-looking methodologies: Review of the various backward-looking methodologies that would be possible providing a summary of pros and cons to be used by SG5 in their assessment by asset class

final alignment 2a.

A short technical analysis of the potential co-existence with the proposed forward-looking methodology - i.e. describing the interaction between forward- and backward-looking methodologies

to start once other tasks completed

2b.

Deliverable Status


3


final alignment 1.


final alignment 2a.



2b.

Deliverable Status

Backward-looking methodologies Overview

Backward-looking methodologies

In arrears In advance Hybrid

Plain / Base Case

Payment delay

Lockout period Lookback Last

reset Last

recent Principal

adjustment Interest rollover

1 2 3 4 5 6 7 8

4

• Clear agreement among Working Group members, that the user guide to overnight risk-free rates(*) published by the Financial Stability Board (FSB) on 4th June 2019, provides a definitive list of backward-looking methodologies.

• Each of these methodologies have been assessed by Sub Group 2A.

(*) Source: https://www.fsb.org/2019/06/overnight-risk-free-rates-a-users-guide/

https://www.fsb.org/2019/06/overnight-risk-free-rates-a-users-guide

Backward-looking methodologies Legend(*):

5

Evaluation Parameters

Observation period Period used to calculate the averaged RFR

Interest period Period for which an interest instalment is paid

…Payment known

…Payment date

2nd Observation period Period used to calculate the adjustment payment

…Adjusted payment date -3M +3M today

1. Operational ease/cash flow management

2. Computational ease/mechanics 3. Hedging ease

5. Period congruency

4. Client acceptance

(*) The graphical descriptions of the backward-looking methodologies refer to the user guide to overnight risk-free rates published by the Financial Stability Board (FSB) on 4th June 2019, p. 8.

Backward-looking methodologies: in arrears 1. Plain / Base Case - Description

Parameter Description Operational ease/cash flow management Operational complexity due to same day payment

Computational ease/mechanics Simple and transparent calculation, rate can be published

Hedging ease Consistent with OIS market, so limited hedging issues

Client acceptance Low due to operational complexity

Period congruency Consistent

Examples of usage Observed in some derivative markets

Conclusion Challenging with a T+1 RFR publication time

• Observation period is identical to the interest period.

• The notional is paid at the start of the period and repaid on the last day of the contract period together with the last interest payment.

• A plain in arrears structure reflects the movement in interest rates over the full interest period and payment is made on the day that it would naturally be due. -3M +3M today

Assessment

6

Backward-looking methodologies: in arrears 2. Payment delay - Description

• Observation period is identical to the interest period.

• Only difference to the plain/base case is the small number of days delay in payment.

Assessment

-3M +3M today

1-5 days of delay of interest payment

7

Parameter Description Operational ease/cash flow management Operational complexity due to small interest payment delay


Hedging ease Consistent with OIS market, so limited hedging issues

Client acceptance High for specific asset classes/users

Period congruency Consistent

Examples of usage OIS derivative market

Conclusion Market standard for many derivatives products, challenging for other users

Backward-looking methodologies: in arrears 3. Lockout period - Description

Parameter Description Operational ease/cash flow management

Sufficient time lag between cash flow fixing and payment date for more sophisticated users

Computational ease/mechanics Slightly higher complexity and lower transparency due to lockout period

Hedging ease Difficult to hedge the lockout period

Client acceptance Potentially limited due to lower transparency – publishing rate difficult

Period congruency Small mismatch

Examples of usage SOFR FRN market

Conclusion Lack of published rate and hedging challenges make it a less viable option

• RFR is no longer updated (i.e. frozen) for a certain number of days prior to the end of the interest period (lockout period).

• The RFR of the prior day to the lockout period is then applied for the lockout period.

-3M +3M today

Assessment

8

Backward-looking methodologies: in arrears 4. Lookback - Description

Parameter Description Operational ease/cash flow management Sufficient time lag between fixing and payment for many users


Hedging ease Easier to hedge than Lockout, but minor risk remain due to small mismatch

Client acceptance High for specific asset classes/users

Period congruency Small mismatch

Examples of usage SONIA FRN market

Conclusion Slightly superior to Lockout approach due to greater hedging and transparency

• Start of observation period a few days prior to the start of interest period.

• Both observation and interest period are of the same length.

-3M +3M today

Assessment

9

Backward-looking methodologies: in advance 5. Last Reset - Description

• Classic fundamentally backward-looking methodology

• Observation period references the previous 3 months to the interest period.

-3M +3M today

Assessment

10

Parameter Description Operational ease/cash flow management Payment rate known in advance, so operationally straightforward


Hedging ease Perfect hedge not available but basis risk

Client acceptance Potentially only workable solution for retail and smaller corporate users

Period congruency Inconsistent

Examples of usage Proposed solution by other working groups for Retail Mortgages

Conclusion Potentially viable option, if rate must be known at the start of the period Potentially very challenging for longer fixing periods, e.g. 12 months

Backward-looking methodologies: in advance 6. Last Recent - Description

Parameter Description Operational ease/cash flow management Payment rate known in advance, so operationally straightforward

Computational ease/mechanics Simple and transparent calculation

Hedging ease No hedging possibility

Client acceptance Difficult due to inconsistency with interest period

Period congruency Inconsistent

Examples of usage Not observed

Conclusion Not an appropriate solution, as it doesn’t correctly reflect the economics or allow for hedging

• Observation period simply references 1 day RFR fixing (or a limited number of days) and this rate is then applied for the whole period.

Assessment

11

-3M +3M today

Backward-looking methodologies: hybrid 7. Principal adjustment - Description

Parameter Description Operational ease/cash flow management Operational complexity due to payment adjustment

Computational ease/mechanics More complex and less transparent calculation

Hedging ease Cash flow mismatch to hedging instrument

Client acceptance Potentially low due to the higher complexity and lower transparency

Period congruency Becomes consistent with the incorporation of the adjustment rate


Conclusion Too high operational complexity vs. benefits

• Incorporates two different observation periods i. Period 1 similar to the in advance approaches

so that rate is known in advance ii. Second observation period used to calculate

adjustment rate – only known at the end of the period

• Adjustment factor used to adjust the payment to equate to average RFR for the current period.

Assessment

-3M +3M today

12

Backward-looking methodologies: hybrid 8. Interest rollover - Description

Parameter Description Operational ease/cash flow management Operational complexity due to payment adjustment

Computational ease/mechanics More complex and less transparent calculation

Hedging ease Cash flow mismatch to hedging instrument

Client acceptance Potentially low due to the higher complexity and lower transparency

Period congruency Becomes consistent with the incorporation of the adjustment rate


Conclusion Too high operational complexity vs. benefits

• Same methodology as the principle adjustment with the only difference that the adjustment payment occurs at some later point beyond current payment date.

Assessment

-3M +3M today

13

Backward-looking methodologies: evaluation overview

14

Methodo-logies

Operational ease / cash flow mgmt.

Computational ease / mechanics Hedging ease Client acceptance Period congruency Examples

of usage Conclusion

1. Plain / Base case

Operational complexity

due to same day payment

Simple and transparent

calculation, rate can be published

Consistent with OIS market so limited hedging issues

Low due to operational complexity

Consistent Observed in some derivative markets

2. Payment delay

Operational complexity due to

small interest payment delay



Consistent with OIS market, so limited

hedging issues

High for specific asset classes/users Consistent OIS derivative

market

3. Lockout period

Sufficient time lag between cash flow fixing and payment

date for more sophisticated users

Slightly higher complexity and

lower transparency due to lockout period

Difficult to hedge the lockout period

Potentially limited due to lower

transparency – publishing rate

difficult

Small mismatch SOFR FRN market

4. Lookback

Sufficient time lag between fixing and payment for many

users



Easier to hedge than Lockout, but minor risk remain due to small mismatch

High for specific asset classes/users Small mismatch SONIA FRN market

5. Last reset Payment rate known

in advance, so operationally

straightforward



Perfect hedge not available but basis

risk

Potentially only workable solution

for retail and smaller corporate

users

Inconsistent Proposed solution by other working groups for Retail Mortgages

/

6. Last recent

Payment rate known in advance, so operationally

straightforward

Simple and transparent calculation

No hedging possibility

Difficult due to inconsistency with

interest period Inconsistent Not observed

7. Principal adjustment


payment adjustment

More complex and less transparent

calculation

Cash flow mismatch to hedging instrument

Potentially low due to the higher

complexity and lower transparency

Becomes consistent with the

incorporation of the adjustment rate

Not observed

8. Interest rollover


payment adjustment

More complex and less transparent

calculation

Cash flow mismatch to hedging instrument

Potentially low due to the higher

complexity and lower transparency

Becomes consistent with the

incorporation of the adjustment rate

Not observed

…viable option …concerns / issues …not viable


15


final alignment 1.


final alignment 2a.



2b.

Deliverable Status

Credit spread methodologies Overview

16

Historic Credit spread methodology

Forward Credit spread methodology

Dynamic Credit spread methodology

High level description

Historical difference between EURIBOR and €STR/EONIA

Simplification of €STR/EURIBOR forward

derivative curve

Replication of the credit/liquidity risk in

another index

Fixed or variable Fixed at point of benchmark cessation

Fixed at point of benchmark cessation

Remains variable over life of the contract

Key issues What historic period to reference?

Reference data at point of cessation

Data collection dependent

Credit spread methodologies 1. Historic Mean/Median Approach – Overview

17

The Historical Mean/Median approach calculates the difference between the realised compounded RFR for a given tenor and the respective EURIBOR fixing for that given tenor/period. Similar to the approach proposed by ISDA.

1. Sensitivity to periods of volatility of the spread

2. How does the spread behave in a hiking/cutting cycle?

3. Simplicity of the method

4. Consistency of term structure, i.e. likelihood of the 12M spread > 6M spread

5. Specificities of the EUR market

Key considerations in the methodological assessment

Credit spread methodologies 1. Historic Mean/Median Approach – Data Analysis (1/2)

18

EUR 6M spread averages for various tenors

• Probability of extreme values clearly higher if averages over a shorter period considered

• Additional high probability of inconsistency of term structure if shorter period considered

• To be noted, that the actual trigger event date far more uncertain than for LIBOR

Credit spread methodologies 1. Historic Mean/Median Approach – Data Analysis (2/2)

19

EUR Mean/Median 5yr average for various tenors

• Clear benefit to the median approach can be seen as a simple way to reduce outliners

• Beyond this, no fundamental concerns with either the mean or median calculation

Credit spread methodologies 1. Historic Mean/Median Approach – Conclusions

20

Simple and transparent if the basic methodology used

Data requirement limited to historical fixings

Preferred method for ISDA for other currencies

Very low probability of manipulation of calculation

Potentially more stable market pricing at point of trigger

× Can be very sensitive to the historic period chosen, this becomes a critical input to the calculation

× May not reflect current market conditions, spot value could therefore be significantly different to calculated spread

× Potential significant value transfer/mark to market impact at point of fallback trigger

PROS CONS

Credit spread methodologies 2. Forward Step Approach – Overview

21

Similar to the ISDA forward approach for the credit spread calculation, the spread adjustment under this approach would be based on observed market levels for the forward EURIBOR/€STR spread in the relevant EURIBOR tenor. However, unlike the original ISDA forward approach proposal which aimed to specify a different spread for every day into the future, this methodology restricts the number of spread values to a limited number of points (in this example 5 tenors). Key reasoning • Simplification • Reliance only on most liquid data points

Credit spread methodologies 2. Forward Step Approach – Overview

t0 = The first calibration date which is the date from which observations of the forward curves commences. It is defined as the date 1Y prior to the discontinuation date.

t1 = The discontinuation date. The first day on which EURIBOR will not be published.

Next, there are the dates which define which credit spread should be applied after cessation. We define the following:

t2 = t1 + 2 years t3 = t1 + 5 years t4 = t1 + 10 years t5 = t1 + 20 years t6 = Potential Final Replacement EURIBOR publication date (t1 + 60Y?)

22

Maturity (Years)

Yield

€STR Swap Curve

EURIBOR Swap Curve

2 5 10

5 Year credit spread



3yr 2yr forward credit spread

calculated

5yr 5yr forward credit spread

calculated

Credit spread methodologies 2. Forward Step Approach – Application

The replacement EURIBOR rate for any given day after cessation will then be obtained by taking the relevant term €STR rate and adding the credit spread according to the following map: For t1 ≤ t < t2 Use 2Y average credit spread For t2 ≤ t < t3 Use 2Y3Y average credit spread For t3 ≤ t < t4 Use 5Y5Y average credit spread For t4 ≤ t < t5 Use 10Y10Y average credit spread For t5 ≤ t < t6 Use 20Y10Y average credit spread

23

Application upon benchmark discontinuation

Term (Years)

Credit spread applied

t1 t2 t3 t4

Average 5yr 5yr forward credit spread applied to benchmark for interest payment in this period

Average 3yr 2yr forward credit spread applied to benchmark for interest payment in this period

Average 2yr forward credit spread applied to benchmark for interest payment in this period

Credit spread methodologies 2. Forward Step Approach – Conclusions

24

More accurate assessment of market value than Historic Mean/Median approach

Value transfer impact should be minimized compared to Historic Mean/Median approach

Only use of liquid market data points enhances transparency and reliability

× Much higher level of complexity than the Historic Mean/Median approach

× Reliant on transparent and stable market for data inputs

× Less accurate than the proposed ISDA methodology

× Whilst significantly reduced, still potentially vulnerable to be influenced by undue transactions/quotes

PROS CONS

Credit spread methodologies

25

• Transaction data of unsecured short-term bank yield products are collected from the primary (term deposits, CP/CD new issues, etc.) and secondary market (CP/CD traded, bonds, etc.).

• Outliers can be cleansed (criteria for the cleansing still need to be defined). • The credit/liquidity spread component can be computed by deducting the term RFR yield from the yield

of the transaction for each available maturity. • Finally a credit/liquidity spread can be calculated for each fixing tenor by interpolating and averaging the

single data points. • ICE Benchmarks Administrators have proposed a similar concept for the U.S. Dollar market (U.S. Dollar

ICE Bank Yield Index).

3. Dynamic Credit spread methodology – Overview & Data Analysis

Credit spread methodologies

26

Key benefit for banks hedging the credit/liquidity risk

In case of a fallback scenario, changes in credit/liquidity spreads of the banking industry over time could be implemented

As the credit/liquidity spread is changing neither quickly nor frequently, transaction data of more than just one day could be used, for example 5-day rolling, 10-day rolling or even longer periods

By aggregating daily transaction data into a rolling period, the volume and diversity of credit data could be improved compared to a daily calculation

× Still faces similar challenge with the underlying transaction data as EURIBOR

× Especially if EURIBOR is discontinued due to lack of available transaction data

× Based on current observations secondary market data might not add that much additional value

× Difficult to source transaction data

× Not considered by trade associations or other working groups

3. Dynamic Credit spread methodology – Conclusions

PROS CONS

2019-08-29 wg on euro rfr meeting item 3 update by ......aug 29, 2019 · sub group 2a: status...

Documents